|
Urban Landuse Change Detection using drone image analysis and hierarchical object-oriented classification
|
Amirhosein Babaeepour   , Asghar Milan * , Saeid Sadeghian |
|
|
|
Abstract: (6 Views) |
With rapid population growth and urbanization, accurate monitoring and evaluation of urban resources through land use/land cover (LULC) change detection has gained increasing significance for effective urban management. This study aims to assess the potential of unmanned aerial vehicle (UAV) imagery for change detection using a rule-based object-oriented classification approach in an urban area. Two datasets were utilized: UAV images captured in 2021 by a SenseFly eBeeX drone over southern Chabahar city, and aerial images acquired in 2013 by an Ultracam XP camera from the same area. A key aspect of the methodology was the implementation of a hierarchical classification model. In this model, all classes were processed using a comprehensive set of features, and a separability matrix was used to design the hierarchical structure. At each node of the hierarchy, optimal features were selected to distinguish between classes. Object-based classification was performed using rule-based analysis of image objects, resulting in the extraction of five land use classes based on statistical thresholds of optimal descriptors. A post-classification processing step was applied to refine the initial results. The quantitative findings showed that the rule-based classification achieved a Kappa coefficient and overall accuracy of 0.90 and 93.31% for the Ultracam imagery, and 0.78 and 83.96% for the UAV orthophoto, respectively. Change detection results revealed that the most significant land cover transformation was from bare land to vegetation (32.843%), while the least change was from buildings to roads (2.34%).
|
Article number: 6 |
| Keywords: Change detection, land use, drone-based photogrammetry, classification, high-resolution images |
|
|
Full-Text [PDF 901 kb]
(3 Downloads)
|
Type of Study: Research |
Subject:
Photo&RS
Received: 2025/05/4
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
